30.51 - 2.14% speedometer Vanilla-ES2015-TodoMVC/DeletingItems/Async / speedometer + 57 more (Android) regression on Sat April 26 2025
Categories
(Firefox for Android :: Crash Reporting, defect)
Tracking
()
Tracking | Status | |
---|---|---|
firefox138 | --- | unaffected |
firefox139 | --- | affected |
firefox140 | --- | affected |
People
(Reporter: intermittent-bug-filer, Unassigned, NeedInfo)
References
(Regression)
Details
(Keywords: perf, perf-alert, regression)
Attachments
(1 obsolete file)
Perfherder has detected a browsertime performance regression from push 73211d3d500c4283ef49da5765a14c1f704cafee. As author of one of the patches included in that push, we need your help to address this regression.
Please acknowledge, and begin investigating this alert within 3 business days, or the patch(es) may be backed out in accordance with our regression policy. Our guide to handling regression bugs has information about how you can proceed with this investigation.
If you have any questions or need any help with the investigation, please reach out to fbilt@mozilla.com. Alternatively, you can find help on Slack by joining #perf-help, and on Matrix you can find help by joining #perftest.
Regressions:
Ratio | Test | Platform | Options | Absolute values (old vs new) | Performance Profiles |
---|---|---|---|---|---|
31% | speedometer Vanilla-ES2015-TodoMVC/DeletingItems/Async | android-hw-a55-14-0-aarch64-shippable | fission webrender | 1.08 -> 1.41 | |
21% | speedometer3 TodoMVC-Lit-Complex-DOM/Adding100Items/Sync | android-hw-a55-14-0-aarch64-shippable | fission webrender | 4.97 -> 6.04 | |
18% | speedometer VueJS-TodoMVC/DeletingAllItems/Async | android-hw-a55-14-0-aarch64-shippable | webrender | 4.24 -> 4.98 | |
15% | speedometer Vanilla-ES2015-Babel-Webpack-TodoMVC/DeletingItems | android-hw-a55-14-0-aarch64-shippable | fission webrender | 13.53 -> 15.57 | |
14% | speedometer Vanilla-ES2015-Babel-Webpack-TodoMVC/DeletingItems/Sync | android-hw-a55-14-0-aarch64-shippable | fission webrender | 12.22 -> 13.90 | |
12% | speedometer Vanilla-ES2015-TodoMVC/DeletingItems | android-hw-a55-14-0-aarch64-shippable | fission webrender | 13.44 -> 15.10 | |
12% | speedometer VueJS-TodoMVC/DeletingAllItems | android-hw-a55-14-0-aarch64-shippable | webrender | 6.97 -> 7.80 | |
12% | speedometer3 TodoMVC-Lit-Complex-DOM/Adding100Items/Sync | android-hw-a55-14-0-aarch64-shippable | webrender | 5.06 -> 5.66 | |
11% | speedometer VueJS-TodoMVC/Adding100Items/Sync | android-hw-a55-14-0-aarch64-shippable | webrender | 4.65 -> 5.18 | |
11% | speedometer Vanilla-ES2015-TodoMVC/DeletingItems/Sync | android-hw-a55-14-0-aarch64-shippable | fission webrender | 12.29 -> 13.66 | |
... | ... | ... | ... | ... | ... |
3% | speedometer Vanilla-ES2015-TodoMVC | android-hw-a55-14-0-aarch64-shippable | webrender | 91.72 -> 94.73 | |
3% | speedometer React-TodoMVC | android-hw-a55-14-0-aarch64-shippable | webrender | 203.10 -> 209.71 | |
3% | speedometer Inferno-TodoMVC/Adding100Items | android-hw-a55-14-0-aarch64-shippable | webrender | 37.39 -> 38.58 | |
3% | speedometer Inferno-TodoMVC | android-hw-a55-14-0-aarch64-shippable | webrender | 211.94 -> 218.20 | |
2% | speedometer | android-hw-a55-14-0-aarch64-shippable | webrender | 116.06 -> 113.58 |
Improvements:
Ratio | Test | Platform | Options | Absolute values (old vs new) | Performance Profiles |
---|---|---|---|---|---|
64% | bing powerUsage | android-hw-a55-14-0-aarch64-shippable | warm webrender | 19,771.84 -> 7,193.82 | |
62% | amazon powerUsage | android-hw-a55-14-0-aarch64-shippable | warm webrender | 20,378.90 -> 7,727.98 | |
62% | youtube-watch powerUsage | android-hw-a55-14-0-aarch64-shippable | fission warm webrender | 17,517.32 -> 6,667.86 | |
62% | wikipedia powerUsage | android-hw-a55-14-0-aarch64-shippable | warm webrender | 17,636.02 -> 6,718.02 | |
62% | bing powerUsage | android-hw-a55-14-0-aarch64-shippable | cold webrender | 19,854.54 -> 7,582.10 | |
... | ... | ... | ... | ... | ... |
2% | dailymail LastVisualChange | android-hw-a55-14-0-aarch64-shippable | warm webrender | 897.60 -> 879.31 |
Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests.
If you need the profiling jobs you can trigger them yourself from treeherder job view or ask fbilt@mozilla.com to do that for you.
You can run all of these tests on try with ./mach try perf --alert 44927
The following documentation link provides more information about this command.
Comment 1•7 days ago
|
||
Set release status flags based on info from the regressing bug 1962204
Comment 2•7 days ago
|
||
The change in bug 1962204 is loading a shared library on the startup path which is probably causing the delay. The power savings on the other hand are obviously caused by the excessive CPU usage going away. I'll see if we have other shared libraries loaded at that point, and if we do if I can fold the native methods into those.
Comment 3•7 days ago
|
||
Mike, I could probably improve this by moving the two native functions I need into mozglue. Would you find that too icky or is it OK?
Comment 4•6 days ago
|
||
Comment 5•6 days ago
|
||
I've uploaded a WIP to give you an idea of what it would look like.
Comment 6•6 days ago
|
||
FWIW if comment 4 only affects startup performance, chances are that speedometer regressions are not because of it. I think it's more likely that the extra CPU usage kept the core in performance mode.
Comment 7•6 days ago
|
||
I agree with what Emilio said. Do we have evidence of a startup regression?
Comment 8•6 days ago
|
||
You're right, I misread the regressions. They're all steady-state ones, there's no startup time regressions. If they're all caused by differences in scheduling then there's nothing I can do here.
Updated•5 days ago
|
Updated•5 days ago
|
Comment 9•12 hours ago
|
||
(In reply to Gabriele Svelto [:gsvelto] from comment #8)
You're right, I misread the regressions. They're all steady-state ones, there's no startup time regressions. If they're all caused by differences in scheduling then there's nothing I can do here.
:gsvelto do you plan on resolving this as wontfix, or will there be any further investigation here?
Description
•